Back to EveryPatent.com
United States Patent |
6,120,459
|
Nitzan
,   et al.
|
September 19, 2000
|
Method and device for arterial blood pressure measurement
Abstract
A method of measurement of arterial diastolic blood pressure includes
generating first and second signals indicative, respectively, of cardiac
induced pulsatile variations in tissue blood volume in a first region and
a second region of the subject's body. These signals are then processed to
derive values of a delay between pulses in the first signal and
corresponding pulses in second signal. A baseline value of the delay is
evaluated, preferably in the absence of externally applied pressure. A
variable pressure is applied to a third region of the subject's body so as
to affect blood flow through at least one artery in the third region, the
variable pressure being varied as a function of time. The first, second
and third regions of the subject's body are chosen such that the delay
varies as a function of the variable pressure. The diastolic pressure is
then identified as a value of the variable pressure corresponding to a
predefined non-zero value of the delay measured relative to the baseline
value.
Inventors:
|
Nitzan; Meir (Beit El DN Mizrach, Binyamin 90631, IL);
Bloch; Louis (9 Briel Beltra, Jerusalem 93695, IL)
|
Appl. No.:
|
328406 |
Filed:
|
June 9, 1999 |
Current U.S. Class: |
600/493; 600/485; 600/490 |
Intern'l Class: |
A61B 005/02 |
Field of Search: |
600/481,485,486,490,493,494,495,500,504
|
References Cited
U.S. Patent Documents
4331155 | May., 1982 | Sacks.
| |
4425920 | Jan., 1984 | Bourland et al.
| |
4437470 | Mar., 1984 | Prost.
| |
4807638 | Feb., 1989 | Sramek | 600/485.
|
4821734 | Apr., 1989 | Koshino.
| |
4860759 | Aug., 1989 | Kahn et al.
| |
5152296 | Oct., 1992 | Simons.
| |
5253645 | Oct., 1993 | Friedman et al. | 600/309.
|
5269310 | Dec., 1993 | Jones et al.
| |
5309908 | May., 1994 | Friedman et al. | 600/694.
|
5423322 | Jun., 1995 | Clark et al.
| |
5564427 | Oct., 1996 | Aso et al. | 600/309.
|
5755669 | May., 1998 | Oho et al. | 600/494.
|
5776071 | Jul., 1998 | Inukai et al. | 600/493.
|
5862805 | Jun., 1999 | Nitzan | 600/479.
|
5865756 | Feb., 1999 | Peel, III | 600/490.
|
Other References
Marmor et al, "Method for Noninvasive Measurement of Central Aortic
Systolic Pressure", Clin. Cardiol., 10: 215-221, 1987.
Geddes et al, "Pulse Arrival Time as a Method of Obtaining Systolic and
Diastolic Blood Pressure Indirectly", Medical & Biological Engineering &
Computing, pp. 671-672, Sep., 1981.
Sharir et al, "Validation of a Method for Noninvasive Measurement Central
Arterial Pressure", Hypertension, 21(1): 74-82, 1993.
|
Primary Examiner: O'Connor; Cary
Assistant Examiner: Natnithithadha; Navin
Attorney, Agent or Firm: Friedman; Mark M.
Claims
What is claimed is:
1. A method for measuring arterial diastolic blood pressure in a subject,
the method comprising:
(a) generating first and second signals indicative, respectively, of
cardiac induced pulsatile variations in tissue blood volume in a first
region and a second region of the subject's body;
(b) processing said first and second signals to derive values of a delay
between pulses in said first signal and corresponding pulses in second
signal;
(c) evaluating a baseline value of said delay;
(d) applying a variable pressure to a third region of the subject's body so
as to affect blood flow through at least one artery in said third region,
said variable pressure being varied as a function of time, said first,
said second and said third regions being chosen such that said delay
varies as a function of said variable pressure; and
(e) identifying a value of said variable pressure corresponding
substantially to a predefined non-zero value of said delay measured
relative to said baseline value.
2. The method of claim 1, wherein said predefined non-zero value of said
delay is between about 30 and about 40 ms.
3. The method of claim 1, wherein said predefined non-zero value of said
delay is between about 15 and about 25 ms.
4. The method of claim 1, wherein said processing includes:
(a) measuring a first amplitude of a first pulse of said first signal;
(b) identifying a first point in the systolic increase portion of said
first pulse at which said first signal reaches a predefined proportion of
said first amplitude;
(c) measuring a second amplitude of a corresponding pulse of said second
signal;
(d) identifying a second point in the systolic increase portion of said
corresponding pulse at which said second signal reaches said predefined
proportion of said second amplitude; and
(e) defining a value of said delay as the time between said first and said
second points.
5. The method of claim 4, wherein said predefined proportion is between
about 0.1 and about 0.5.
6. The method of claim 1, wherein said processing includes:
(a) calculating a time derivative of a first pulse of said first signal;
(b) identifying a first maximum value of said time derivative in the
systolic increase portion of said first pulse;
(c) identifying a first point at which said time derivative reaches a
predefined proportion of said first maximum value;
(d) calculating a time derivative of a second pulse of said second signal;
(e) identifying a second maximum value of said time derivative in the
systolic increase portion of said second pulse;
(f) identifying a second point at which said time derivative reaches a
predefined proportion of said second maximum value; and
(g) defining a value of said delay as the time between said first and said
second points.
7. The method of claim 1, wherein said processing includes:
(a) for corresponding pulses of each of said first and second signals,
(i) identifying a local minimum of said signal,
(ii) fitting a negative gradient line to a predefined portion of said
signal prior to said local minimum,
(iii) fitting a positive gradient line to a predefined portion of said
signal subsequent to said local minimum, and
(iv) extrapolating said negative gradient line and said positive gradient
line to determine an intersection, referred to as an adjusted minimum
point of said signal; and
(b) defining a value of said delay as the time between said adjusted
minimum point of said first signal and said adjusted minimum point of said
second signal.
8. The method of claim 1, wherein said processing includes:
(a) evaluating a measure of correlation between corresponding pulses of
said first and second signals, said measure of correlation being evaluated
as a function of a time shift of said second signal relative to said first
signal; and
(b) defining a value of said delay as the time shift which generates a
maximum value of said measure of correlation.
9. The method of claim 1, wherein said second region is chosen such that
variations in said subcutaneous blood volume in said second region are
substantially unaffected by variations in said variable pressure.
10. The method of claim 1, wherein said first and second signals are
generated by use of non-invasive sensors.
11. The method of claim 1, wherein said first and second signals are
generated by use of photoplethysmography sensors.
12. The method of claim 8, wherein said variable pressure is applied using
an inflatable cuff, and wherein at least one of said photoplethysmography
sensors is attached to said cuff.
13. The method of claim 1, wherein said first and second signals are
generated by use of impedance plethysmography sensors.
14. A device for measuring arterial diastolic blood pressure in a subject,
the device comprising:
(a) a pressure cuff applicable to a first region of the subject's body so
as to affect blood flow through at least one artery in said first region;
(b) a pressure controller operatively connected to said pressure cuff so as
to vary a current pressure of said pressure cuff;
(c) first and second plethysmography sensors for application to a second
region and a third region of the subject's body, said first and second
plethysmography sensors being configured to produce first and second
signals, respectively, indicative of pulsatile variations in tissue blood
volume in said second and third regions, respectively; and
(d) a processor associated with said pressure controller and with said
first and second plethysmography sensors, said processor being configured
to:
(i) process said first and second signals to derive values of a delay
between pulses in said first signal and corresponding pulses in second
signal,
(ii) evaluate a baseline value of said delay corresponding to a current
pressure substantially equal to ambient pressure, and
(iii) identify as the diastolic pressure a value of said variable pressure
corresponding substantially to a predefined non-zero value of said delay
measured relative to said baseline value.
15. A method for measuring arterial diastolic blood pressure in a subject,
the method comprising:
(a) generating first and second signals indicative, respectively, of
cardiac induced pulsatile variations in tissue blood volume in a first
region and a second region of the subject's body;
(b) processing said first and second signals to derive values of a measure
of correlation between pulses in said first signal and corresponding
pulses in second signal;
(c) applying a variable pressure to a third region of the subject's body so
as to affect blood flow through at least one artery in said third region,
said variable pressure being varied as a function of time, said first,
said second and said third regions being chosen such that said measure of
correlation varies as a function of said variable pressure; and
(d) identifying a value of said variable pressure corresponding to a value
of said measure of correlation substantially equal to a predefined value.
16. The method of claim 15, further comprising evaluating a baseline value
of said measure of correlation, said measure of correlation being adjusted
on the basis of said baseline value such that said measure of correlation
approaches 1 at low values of said variable pressure.
17. The method of claim 15, wherein said predefined value is at least about
0.9.
Description
FIELD AND BACKGROUND OF THE INVENTION
The present invention relates to a method and device for the measurement of
arterial blood pressure (ABP). More particularly, the present invention
relates to a method and a device for measuring systolic and diastolic
blood pressure.
Blood supply to the tissues of a living body is essential for maintaining
their metabolism and proper function. During systole (heart contraction),
blood is ejected from the heart into the arterial system, thereby
increasing the arterial blood pressure. The maximal arterial blood
pressure (ABP) is the systolic blood pressure (SBP). During and after
systole, blood flows from the arteries, through the capillaries, into the
veins, and from them back into the heart. The period between two systoles
is called diastole. During diastole, the arterial blood pressure
decreases; the minimal arterial blood pressure (at the end of diastole) is
called diastolic blood pressure (DBP).
Similar to blood pressure, blood volume in the tissue also shows
oscillations at the heart rate. During systole, blood is ejected from the
left ventricle into the peripheral tissues, thereby increasing their blood
content. The measurement of the cardiac induced changes of tissue blood
volume is called plethysmography, which can be performed by means of
several methods, including photoplethysmography (PPG), which is the
measurement of light absorption in tissue. The PPG signal originates from
the increase of tissue blood volume during systole, and the consequent
higher light absorption. FIG. 1 shows a known PPG probe attached to a
finger. The light source L emits light into the tissue and the
photodetector D measures the light scattered from the tissue under the
skin. The output of the photodetector depends on the tissue blood volume,
and oscillates with the oscillations of the latter.
FIG. 2 shows the blood pressure and the PPG signal measured simultaneously
in the finger arteries as a function of time. The blood pressure
measurement was performed on a fingertip by means of a continuous,
non-invasive blood pressure meter (Finapres, Ohmeda, U.S.A.). As can be
seen in FIG. 2, the curve of oscillations (at the heart rate) of the
tissue blood volume as measured by the PPG signal, is similar, but not
identical, to the ABP curve.
Blood pressure can change because of exercise, mental stress, or
excitement. It also changes spontaneously due to activity of the autonomic
nervous system. For adults aged below 40 years, the values of normal blood
pressure (at rest) are 120 mmHg and 80 mmHg for systolic and diastolic
blood pressures, respectively; If the ABP is too high (hypertension), the
subject is at higher risk of cerebral stroke and heart attack. Lower than
normal blood pressure (hypotension) is acutely hazardous, since it may
cause low blood supply to the brain, resulting in fainting or even in
brain damage. Decreasing blood pressure for patients after trauma, surgery
or heart attack is an indication of cardiovascular deterioration.
Blood pressure can be measured invasively by inserting a catheter into an
artery and measuring the pressure by means of a piezoelectric device. This
measurement is the most reliable one, and it is done in intensive care
units where an arterial line is inserted for additional purposes. Due to
its invasiveness, this method is not used for routine applications.
The auscultatory method is the most common method for non-invasive
measurement of blood pressure, and is based on hearing (via stethoscope or
microphone) the turbulence sounds which appear in a compressed artery when
it is intermittently closed and opened by means of an inflatable cuff
having air pressure of a value between that of diastolic and systolic
blood pressure. Usually, the cuff air pressure is increased above the SBP,
then decreased. The cuff air pressure at which the turbulence sounds
appear is the SBP; the pressure at which the quality of the sounds
changes, becoming muffled, is defined as the higher DBP (IV Korotkoff or
phase IV DBP); the pressure at which they totally disappear is defined as
the lower DBP (V Korotkoff or phase V DBP). In general the lower DBP has
to be taken as the DBP, but for some groups of patients for whom the
Korotkoff sounds are heard even for extremely low cuff pressure, such as
in pregnant women, the higher DBP is taken. The manual auscultatory method
(using a stethoscope) has been accepted as the gold standard for
non-invasive ABP measurement, and is routinely used in clinics and
hospitals. The automatic auscultatory method (using a microphone), is also
used for monitoring ABP in hospital wards. Despite its extensive use, the
auscultatory method is not accurate, both because of the difficulty in
detecting the correct sounds and because of the unclear relationship
between the disappearance of the turbulence sounds and DBP.
Automatic blood pressure measurement can also be done by means of the
oscillometric method. A cuff is applied to the arm or finger and, besides
the measurement of the average air pressure, the oscillatory variations of
air pressure in the cuff are measured by means of a piezoelectric pressure
transducer. Oscillations at the rate of the heart can then be seen in the
cuff pressure (oscillometry) due to the cardiac induced changes in the
arterial blood volume. In an alternative method, a sensor for detecting
blood volume changes in the arteries, such as a PPG device, is attached to
the skin under the cuff. Here too, oscillations at the rate of the heart
appear in the volume sensor output (volume oscillometry). When the air
pressure is continuously increased above diastolic blood pressure, these
oscillations also increase until the air pressure is equal to the mean
blood pressure, and then they decrease. The systolic and diastolic blood
pressure can be derived from the curve of the amplitude of oscillation as
a function of the air pressure, using empirical formulae. This method,
which is called "oscillometry", can be used for monitoring blood pressure,
but the measurement time is long: more than 20 heart beats, depending on
the patient and on the required accuracy. In any case, the method and the
commercial devices which are based thereon are not considered to be
accurate.
The low accuracy of the automatic auscultatory and oscillometry methods for
the measurement of diastolic and systolic blood pressure, and the need for
a reliable automatic method, have resulted in several attempts to develop
other methods for blood pressure measurements. Some of these methods are
based on PPG measurement. The systolic blood pressure can be
non-invasively measured by means of PPG, by using a PPG device and a cuff
around the arm or finger, increasing the air pressure in the cuff, and
determining the air pressure at which the PPG signal disappears. This air
pressure is equal to the systolic blood pressure in the artery under the
cuff. In principle, measurement of systolic blood pressure by PPG and a
pressure cuff may be performed in a straightforward manner by identifying
the onset of PPG pulses. Determination of the diastolic blood pressure
from the PPG signal, on the other hand, is more difficult.
In U.S. Pat. No. 5,269,310, there is disclosed a method for measuring, by
means of PPG, changes of blood volume in the arteries during systole
together with the patient's blood pressure, and for determining what is
assumed to be a constant k particular to the patient's arterial blood
pressure-volume relationship. By means of this calibration, the DBP and
SBP for each heartbeat is determined from the minimum and maximum points
of the PPG signal. The method is not accurate, since DBP and SBP are not
actually related to the maximum and minimum of the PPG signal by a
constant k.
In U.S. Pat. No. 5,423,322, an exponential relationship is assumed between
the ABP and the blood volume changes measured by PPG, for the assessment
of the cardiac-induced blood pressure oscillations from the simultaneous
blood volume oscillations in the heart rate. There are several drawbacks
to this method, as will now be detailed.
Firstly, the relationship between the arterial blood pressure and the blood
volume is not strictly exponential. In fact, the volume vs. pressure curve
changes as a function of time, and even changes between the period of
increasing pressure (systole) to the period of decreasing pressure
(diastole) within the same cardiac cycle, as can be seen in FIG. 2 of the
present application.
Secondly, the blood volume changes not only in a single artery, but also in
the small arteries and in the arterioles (resistance vessels). It is not
possible to simulate the entire group of arteries and arterioles as a
single artery, since the pressure therein is not constant due to the
reduction of the blood pressure from the arteries to the arterioles.
Another known method for continuous measurement of finger ABP is the
arterial volume clamp method, which is based on PPG. The device utilized
for this method is composed of a finger cuff with a PPG probe, and the
method is based on the determination of the cuff air pressure which is
required to keep the arterial blood volume constant. The device enables
the measurement of ABP changes during the cardiac cycle via very rapid
changes of the cuff air pressure. The method is very sophisticated, but it
was not found to reliably record ABP. The device is expensive, due to the
need to swiftly change the cuff air pressure in accordance with the blood
pressure changes during the cardiac cycle.
Other methods for the measurement of ABP have been suggested, but the only
methods which have been accepted for routine and comprehensive clinical
use are the oscillometric and auscultatory methods, indicating that the
other suggested methods are either not reliable enough, or are too
complicated, for clinical use.
Another approach, suggested by a number of academic papers but not
implemented in practice, proposes to measure diastolic blood pressure on
the basis of a delay in the pulse caused by pressure from a cuff. Applying
a pressure between diastolic and systolic blood pressure on an artery
results in compression of the artery for part of the cardiac cycle time as
can be seen in FIG. 3. As a result, the pressure pulse in the artery
distal to the pressure application location will start later than in
contralateral arteries not affected by the pressure. This approach was
presented by L. A. Geddes et al. in a paper entitled "Pulse Arrival Time
as a Method of Obtaining Systolic and Diastolic Blood Pressure
Indirectly", (Medical & Biological Engineering & Computing, September
1981, 19:pp. 671-672). Geddes et al., experimenting on dogs, compared the
measurements of an invasive pressure sensor in the leg of a dog with
either another similar sensor in the contralateral leg or an ECG reference
to detect a delay in the pulse reaching a location beyond a pressure cuff.
The use of ECG as a time reference is particularly problematic, giving
broad scattering of results. The measurement of the pulse delay due to the
cuff pressure using ECG as a reference was also suggested by A. Marmor et
al. (Clin. Cardiol. 1987, 10:215-221) and T. Sharir et al. (Hypertension
1993, 21: 74-82) for the determination of the systolic increase curve of
arterial pressure as a function of time.
Even with a second sensor in the contralateral leg as a reference, the
point at which the diastolic pressure is supposedly indicated appears
poorly defined. Furthermore, the measurement of the time delay between the
pulses in the two sides is inaccurate, since the time delay was identified
as the time difference between the minima of the corresponding pressure
pulses in the two sides. The measurement of the time of the pulse minimum
is subject to significant error, since the curve in the neighborhood of
the minimum changes slowly and a small error in the pressure measurement
may result in a large error in the determination of the minimum time. This
problem would be accentuated if the noninvasive method were used for the
determination of the start of the pulse, in a peripheral region, since
noninvasive measurement of any parameter which is related to the pulse
pressure has a higher noise level than direct invasive measurement of
arterial blood pressure. It should be noted that Geddes et al. claimed
that it is possible to use a noninvasive technique to detect the start of
the pressure pulse, and that they intend to do that, and to publish the
results in a second paper. However, to the best of our knowledge, no such
paper has ever been published. It seems that the analysis of the
non-invasively achieved signal as suggested by Geddes did not permit
accurate measurement of the diastolic blood pressure, possibly for the
reasons discussed above.
There is therefore a need for a practical, non-invasive technique for
measuring arterial diastolic blood pressure on the basis of a delay in
pulses caused by a pressure cuff. It would also be highly advantageous to
provide an automatic device for measuring arterial diastolic blood
pressure according to such a technique.
SUMMARY OF THE INVENTION
The present invention is device and method for measurement of arterial
blood pressure and, in particular, the diastolic blood pressure.
According to the teachings of the present invention there is provided, a
method for measuring arterial diastolic blood pressure in a subject, the
method comprising: (a) generating first and second signals indicative,
respectively, of cardiac induced pulsatile variations in tissue blood
volume in a first region and a second region of the subject's body; (b)
processing the first and second signals to derive values of a delay
between pulses in the first signal and corresponding pulses in second
signal; (c) evaluating a baseline value of the delay; (d) applying a
variable pressure to a third region of the subject's body so as to affect
blood flow through at least one artery in the third region, the variable
pressure being varied as a function of time, the first, the second and the
third regions being chosen such that the delay varies as a function of the
variable pressure; and (e) identifying a value of the variable pressure
corresponding substantially to a predefined non-zero value of the delay
measured relative to the baseline value.
In preferred cases, the phase IV DBP is identified directly by use of a
value of the delay between about 30 and about 40 ms, and the phase V DBP
is identified directly by use of a value of the delay between about 15 and
about 25 ms.
According to a first preferred approach, the processing includes: (a)
measuring a first amplitude of a first pulse of the first signal; (b)
identifying a first point in the systolic increase portion of the first
pulse at which the first signal reaches a predefined proportion of the
first amplitude; (c) measuring a second amplitude of a corresponding pulse
of the second signal; (d) identifying a second point in the systolic
increase portion of the corresponding pulse at which the second signal
reaches the predefined proportion of the second amplitude; and (e)
defining a value of the delay as the time between the first and the second
points.
According to a second preferred approach, the processing includes: (a)
calculating a time derivative of a first pulse of the first signal; (b)
identifying a first maximum value of the time derivative in the systolic
increase portion of the first pulse; (c) identifying a first point at
which the time derivative reaches a predefined proportion of the first
maximum value; (d) calculating a time derivative of a second pulse of the
second signal; (e) identifying a second maximum value of the time
derivative in the systolic increase portion of the second pulse; (f)
identifying a second point at which the time derivative reaches a
predefined proportion of the second maximum value; and (g) defining a
value of the delay as the time between the first and the second points.
According to a third preferred approach, the processing includes: (a) for
corresponding pulses of each of the first and second signals, (i)
identifying a local minimum of the signal, (ii) fitting a negative
gradient line to a predefined portion of the signal prior to the local
minimum, (iii) fitting a positive gradient line to a predefined portion of
the signal subsequent to the local minimum, and (iv) extrapolating the
negative gradient line and the positive gradient line to determine an
intersection, referred to as an adjusted minimum point of the signal; and
(b) defining a value of the delay as the time between the adjusted minimum
point of the first signal and the adjusted minimum point of the second
signal.
According to a fourth preferred approach, the processing includes: (a)
evaluating a measure of correlation between corresponding pulses of the
first and second signals, the measure of correlation being evaluated as a
function of a time shift of the second signal relative to the first
signal; and (b) defining a value of the delay as the time shift which
generates a maximum value of the measure of correlation.
According to a further feature of the present invention, the first and
second signals are generated by use of non-invasive sensors, and
preferably, photoplethysmography sensors.
There is also provided according to the teachings of the present invention,
a device for measuring arterial diastolic blood pressure in a subject, the
device comprising: (a) a pressure cuff applicable to a first region of the
subject's body so as to affect blood flow through at least one artery in
the first region; (b) a pressure controller operatively connected to the
pressure cuff so as to vary a current pressure of the pressure cuff; (c)
first and second plethysmography sensors for application to a second
region and a third region of the subject's body, the first and second
plethysmography sensors being configured to produce first and second
signals, respectively, indicative of pulsatile variations in tissue blood
volume in the second and third regions, respectively; and (d) a processor
associated with the pressure controller and with the first and second
plethysmography sensors, the processor being configured to: (i) process
the first and second signals to derive values of a delay between pulses in
the first signal and corresponding pulses in second signal, (ii) evaluate
a baseline value of the delay corresponding to a current pressure
substantially equal to ambient pressure, and (iii) identify as the
diastolic pressure a value of the variable pressure corresponding
substantially to a predefined non-zero value of the delay measured
relative to the baseline value.
There is also provided according to the teachings of the present invention,
a method for measuring arterial diastolic blood pressure in a subject, the
method comprising: (a) generating first and second signals indicative,
respectively, of cardiac induced pulsatile variations in tissue blood
volume in a first region and a second region of the subject's body; (b)
processing the first and second signals to derive values of a measure of
correlation between pulses in the first signal and corresponding pulses in
second signal; (c) applying a variable pressure to a third region of the
subject's body so as to affect blood flow through at least one artery in
the third region, the variable pressure being varied as a function of
time, the first, the second and the third regions being chosen such that
the measure of correlation varies as a function of the variable pressure;
and (d) identifying a value of the variable pressure corresponding to a
value of the measure of correlation substantially equal to a predefined
value.
According to a further feature of the present invention, a baseline value
of the measure of correlation is evaluated, the measure of correlation
being adjusted on the basis of the baseline value such that the measure of
correlation approaches 1 at low values of the variable pressure.
According to a further feature of the present invention, the predefined
value is at least about 0.9.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention will now be described in connection with certain preferred
embodiments with reference to the following illustrative figures so that
it may be more fully understood.
With specific reference now to the figures in detail, it is stressed that
the particulars shown are by way of example and for purposes of
illustrative discussion of the preferred embodiments of the present
invention only, and are presented in the cause of providing what is
believed to be the most useful and readily understood description of the
principles and conceptual aspects of the invention. In this regard, no
attempt is made to show structural details of the invention in more detail
than is necessary for a fundamental understanding of the invention, the
description taken with the drawings making apparent to those skilled in
the art how the several forms of the invention may be embodied in
practice.
In the drawings:
FIG. 1 illustrates a prior art PPG probe attached to a finger of a subject;
FIG. 2 shows curves of the cardiac-induced oscillations of the blood
pressure in a finger's arteries and the cardiac-induced oscillations of
the tissue blood volume as a function of time;
FIG. 3 is a plot of arterial blood pressure as a function of time;
FIG. 4 is an illustration of an embodiment of the device for diastolic
blood pressure measurement according to the present invention;
FIG. 5 is a block diagram of the device of FIG. 4;
FIG. 6 shows curves of PPG signals in the fingers of the right and left
hands of a subject, for different air pressures applied to the subject's
right arm;
FIG. 7 illustrates a further embodiment of the device for diastolic blood
pressure measurement;
FIG. 8 illustrates a still further embodiment of the device for diastolic
blood pressure measurements;
FIG. 9 is a cross-sectional view of the pressure application means and the
PPG probe of FIG. 8;
FIG. 10 illustrates a first example of a technique for measuring time delay
between corresponding pulses of two PPG signals according to the present
invention;
FIG. 11 illustrates a second example of a technique for measuring time
delay between corresponding pulses of two PPG signals according to the
present invention;
FIG. 12 illustrates a third example of a technique for measuring time delay
between corresponding pulses of two PPG signals according to the present
invention;
FIG. 13 illustrates results of time delay measurements as a function of
applied cuff pressure obtained by the technique of FIG. 10;
FIG. 14 illustrates results of time delay measurements as a function of
applied cuff pressure obtained by the technique of FIG. 11;
FIG. 15 illustrates a correlation function taken between two PPG signals as
a function of applied cuff pressure;
FIG. 16 shows variations in both applied cuff pressure and an affected PPG
signal over a given time period;
FIGS. 17A, 17B and 17C show comparisons of arterial blood pressure
measurements obtained by the methods of the present invention and by
conventional sphygmomanometry for systolic, first diastolic and second
diastolic blood pressures, respectively;
FIG. 18 shows a first implementation of the device of FIG. 4 for use in
blood pressure monitoring; and
FIG. 19 shows a second implementation of the device of FIG. 4 for use in
blood pressure monitoring.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention is a device and method for measurement of arterial
blood pressure and, in particular, the diastolic blood pressure.
The principles and operation of devices and methods according to the
present invention may be better understood with reference to the drawings
and the accompanying description.
Referring now to the drawings, FIGS. 4, 5 and 7-9 show various embodiments
of a device, constructed and operative according to the teachings of the
present invention, for measuring the arterial blood pressure of a subject.
Generally speaking, the invention is based on the application of a pressure
cuff with air pressure Pa around the limb of a subject on a pressure
application site, and a first PPG probe on a measurement site, distal to
the pressure application site, said first PPG probe producing the first
PPG signal. When the cuff is inflated to an air pressure above the SBP,
the arteries under the cuff will be compressed and closed. Hence, the SBP
can be measured by determining the air pressure at which the PPG signal
disappears, as described above. When the cuff air pressure is held at a
given value of Pa.sub.0, which is below the SBP but above the DBP (see
FIG. 3), the arteries under the cuff are compressed and closed for those
short periods of time (between t.sub.1 and t.sub.2) in which the arterial
blood pressure is below the cuff air pressure Pa.sub.0. During the time at
which the blood pressure decreases below the applied air pressure
Pa.sub.0, the blood stops flowing through the momentarily closed artery
and no systolic increase in the tissue blood volume occurs, resulting in a
reduction in said first PPG signal amplitude and a delay in the start of
the systolic increase in said first PPG signal. Both the change in the PPG
amplitude and the delay of the systolic increase are not easy to detect,
since the PPG signal changes spontaneously. This small change in the first
PPG signal can, however, be detected by measuring the first PPG signal
simultaneously with a second (reference) PPG signal in a second PPG probe,
which is not placed distal to the cuff so that it is not directly affected
by the cuff, and comparing the two PPG signals. When the air pressure
reaches a value between that of the diastolic and systolic blood pressure,
the first PPG signal in the site distal to the cuff is delayed relative to
the second (reference) signal of the second PPG probe, the length of the
time delay increasing as the cuff pressure increases.
As mentioned above in the context of the work of Geddes et al.,
determination of the diastolic blood pressure from measurements of the
delay between two such signals is non-trivial. By providing a combination
of initial baseline time-delay determination, effective algorithms for
evaluating the delay, and parameters determined by extensive
experimentation, most preferred implementations of the present invention
offer an accurate, repeatable and experimentally verified device and
technique for measuring diastolic blood pressure, as will be described.
Thus, the method of the present invention includes generating first and
second signals indicative, respectively, of cardiac induced pulsatile
variations in tissue blood volume in a first region and a second region of
the subject's body. These signals are then processed to derive values of a
time-delay between pulses in the first signal and corresponding pulses in
second signal. The delay may be calculated by application of one or more
algorithm, selected examples of which will be described below. A baseline
value of the delay is evaluated in the absence of externally applied
pressure. A variable pressure is applied to a third region of the
subject's body so as to affect blood flow through at least one artery in
the third region, the variable pressure being varied as a function of
time. The first, second and third regions of the subject's body are chosen
such that the delay between the first and second signals varies as a
function of the variable pressure applied to the third region. The
diastolic pressure is then identified as a value of the variable pressure
corresponding substantially to a predefined non-zero value of the delay
measured relative to the baseline value.
In structural terms, with reference to FIG. 4, the invention provides a
device for measuring arterial blood pressure in a subject, including a
pressure cuff 18 applicable to a first region of the subject's body so as
to affect blood flow through at least one artery in the first region, and
a pressure controller 12a operatively connected to the pressure cuff so as
to vary a current pressure of pressure cuff 18. First and second
plethysmography sensors 2 and 4, applicable to two regions of the
subject's body, are configured to produce first and second signals,
respectively, indicative of systolic pulsatile variations in tissue blood
volume in the corresponding regions. A processor 12b is associated with
pressure controller 12a, optionally in a single unit 12, and with first
and second plethysmography sensors 2 and 4. Processor 12b is configured
to: (i) process the first and second signals to derive values of a delay
between pulses in the first signal and corresponding pulses in second
signal, (ii) evaluate a baseline value of the delay corresponding to a
current pressure substantially equal to ambient pressure, (iii) identify a
current value of the variable pressure corresponding to each value of the
delay, and (iv) identify as the diastolic pressure a value of the variable
pressure corresponding substantially to a predefined non-zero value of the
delay measured relative to the baseline value.
It will be noted that sensors 2 and 4 may be applied to any two regions of
the subject's body chosen such that the time delay obtained varies as a
function of the pressure applied to cuff 18. In the case of FIG. 4, the
reference signal from sensor 4 is substantially unaffected by variations
in the applied pressure. In alternative implementations such as that of
FIG. 7, the two regions in which measurements are taken may both be
affected so long as they are affected to a different extent, thereby also
allowing measurement of a time delay which varies as a function of the
applied pressure.
Turning now to the embodiment of FIGS. 4 and 5 in more detail, the device
consists of a first PPG probe 2, fitted with per se known means for
attaching the probe to a finger of one hand of a subject and similarly, a
second PPG probe 4 fitted with means for attaching it to a finger of the
second hand of said subject. The probes 2 and 4 each include a light
source L modulated by a modulator 6 and a photodetector D, the output of
which is advantageously amplified, filtered and demodulated at 8 before
being digitized by an A/D converter 10. The outputs for converter 10 are
applied to a processor/controller 12. The latter also governs the
operation of pump 16, which affects the inflation and deflation of a
pressure application means 18, e.g., a cuff, configured to be attached to
the arm of one of the subject's hands and receives information from an air
pressure monitor 14. While the modulator 6, amplifiers/filters 8 and A/D
converters 10 are shown for the sake of clarity as being a separate
assembly, it should be understood that in practice, these functions are
performed by circuits physically constituting parts of the
processor/controller 12. The device also includes a display 20 for
displaying the arterial blood pressure and other selectable, useful
information.
FIG. 6 shows the PPG signals in the fingers of both the right and left
hands simultaneously measured when a pressure application means 18 is
attached to the right arm and the air pressure in it is between the
systolic and diastolic blood pressures. In the right hand curve a delayed
onset of pulses as well as a reduced amplitude and area of pulses of the
PPG signal when the external air pressure exceeds the diastolic ABP can be
clearly seen.
As explained above, the basis for the invention is to compare the onset
time of the first PPG signal in the first site which is distal to the
pressure application means with that of another reference signal in the
second site which is not affected by the pressure of means 18, or which is
affected in a different manner from the signal of the first site. Hence,
the measurement of the diastolic blood pressure can also be obtained by
changing the pressure in the pressure application means and measuring the
resultant delay between the first and the second PPG signals by using two
PPG probes on the same limb, one distal to means 18 and the other proximal
to it.
Accordingly, and with reference to FIG. 7, there is shown an embodiment
similar to the embodiment of FIG. 4, however, in which the PPG probes 2
and 4 are attached to the subject's arm on both edges of the pressure
application means 18, instead of on the fingers of both hands, as seen in
FIG. 3. Thus, a combined pressure application means 18 and PPG probes may
be formed, which can be affixed at any suitable location on a subject's
body in such a way that one of the PPG probes, e.g., probe 2, will be
located to provide measurements of light absorption in tissue distal to
the pressure application means 18 relative to the heart, while the other
PPG probe 4 will be located to provide measurements of light absorption in
tissue proximal thereof.
FIGS. 8 and 9 illustrate another embodiment of the device in accordance
with the present invention, in which the PPG probe 2, which measures the
blood volume changes which are affected by the pressure application means
18, is attached to the skin of the forearm under the pressure application
means. The second probe 4, which measures the blood volume changes which
are not affected by the pressure application means 18, is attached to the
skin of the contralateral forearm. The signals of the two probes are
compared, and the time delay is analyzed as described herein.
While the above preferred embodiments specifically utilize the measurement
of the systolic increase of tissue blood volume by the PPG probes, it
should be noted that other plethysmography sensors such as electrical
impedance plethysmography sensors may equally be used. Furthermore, it
should be understood that non-invasive detectors of another cardiovascular
parameter, such as blood flow, blood velocity or tissue blood pressure,
could be utilized.
Parenthetically, it should also be noted that, while reference is made
herein to preferred examples in which measurements are made while
decreasing the applied air pressure, alternative embodiments may be
implemented in which measurements are taken while increasing the pressure.
As mentioned above, evaluation of the time-delay between corresponding
pulses of the two signals is non-trivial and, particularly in the case of
noisy outputs from non-invasive sensors, cannot be performed by the
simplistic approach presented by the Geddes et al. reference mentioned
above. In order to provide a practical and accurate device and method, the
present invention preferably performs preprocessing in the form of various
smoothing functions followed by one or more of a number of techniques for
measuring the time delay, examples of which will now be described.
For conciseness of presentation, each technique will be described here in
algorithmic terms only. The practical details of an implementation, as
well as the hardware required for implementing processor 12b, will be
clear to one ordinarily skilled in the art from the algorithms described.
Typically, a microprocessor unit is employed operating conventional
computational software under a suitable operating system. Alternatively, a
custom hardware implementation, or a combination of hardware and software
(referred to as "firmware") may be used.
Turning first to the preprocessing, problems of noise common in PPG signals
are preferably minimized by one or more smoothing function. In a first
preferred embodiment, initial smoothing is achieved by employing a high
initial sampling rate followed by averaging of groups of readings to
obtain a lower effective sampling rate. In a typical example, readings may
be taken at a rate of 5 kHz followed by averaging of groups of 5 readings,
giving an effective sampling rate of 1 kHz. Additionally, or
alternatively, smoothing algorithms are applied which do not alter the
effective sampling rate. A simple example of such an algorithm is
averaging over a sliding window. Thus in the typical example mentioned
above, each of the 1000 readings per second may be set to a value
corresponding to an average of itself together with about 20 preceding
values and about 20 subsequent values.
The overall effect of such preprocessing is preferably to remove, or
minimize the effect of, any noise or other transient features which have
frequencies one or more orders of magnitude higher than those of the pulse
rate. Thus, in preferred implementations, the preprocessing removes or
attenuates at least those components of the signals with frequencies in
excess of about 100 Hz, and preferably those with frequencies in excess of
about 25 Hz.
Turning now to specific examples of techniques for evaluating the delay, it
is a preferred feature of the present invention that the local minimum is
not relied upon as a reference point to evaluate the time delay, thereby
avoiding a substantial cause of inaccuracy in the approach presented by
Geddes et al. Instead, the present invention either employs an alternative
reference point such as will be described with reference to FIGS. 10-12,
or an alternative technique such as will be described thereafter.
A first preferred approach, illustrated in FIG. 10, entails measuring the
amplitude A.sub.1 of a pulse of the first signal and identifying a first
point in the systolic increase portion of the pulse at which the first
signal reaches a predefined proportion, in this case 0.1, of the amplitude
A.sub.1. The same is done for the corresponding pulse of the second
signal, thereby identifying a point in the systolic increase portion of
the corresponding pulse at which the second signal reaches the same
predefined proportion of an amplitude A.sub.2, where A.sub.2 is the
amplitude of the corresponding pulse of the second signal. The time delay
.DELTA.t is then identified as the time between these two points.
A second preferred approach, illustrated in FIG. 11, employs taking the
derivative of the systolic increase portion of each PPG curve and finding
the maximum values d1.sub.max and d2.sub.max of the derivative for each.
The point on each derivative curve at which the derivative reaches a
predetermined proportion, in this case 50%, of its maximum value is then
taken as a reference point for determining the delay. In the example shown
here, these reference points are identified as d1.sub.max/2 and
d2.sub.max/2.
A third preferred approach, illustrated in FIG. 12, employs a geometrical
construct to generate a well-defined reference point in the region of the
local minimum of the signals. This approach may be regarded as sharpening
the minimum to a point to obtain a well defined adjusted minimum point.
One particularly straightforward and advantageous implementation of this
approach is to fit a straight line, by least squares approximations or
other techniques, to portions of the signal slightly before and after the
minimum. In one preferred implementation which has been found to give good
results, straight lines were fitted to portions of the signals between
about 20-50 ms before and after the minimum.
As mentioned above, the present invention also provides alternative
techniques which avoid the need to define a specific reference point in
the signals. In a primary example, this is done by evaluating a
correlation coefficient employing the entirety of the pulse signals.
Specifically, a measure of correlation between corresponding pulses of the
first and second signals is evaluated as a function of a time shift of the
second signal relative to the first signal. The time shift which generates
a maximum value of the measure of correlation, is then defined as the
delay between the pulses.
By way of example, the correlation coefficient for two sampled functions X
and Y with lag .tau. between them may be calculated by comparing the two
series x(n) and y(n), (n=1, 2, 3, . . . N, where N is the number of the
PPG pulses) after removing the last .tau. terms of the x(n) series and the
first .tau. terms of the y(n) series. Hence the two series to be compared
are:
x(1), x(2), . . . x(N-.tau.) and y(.tau.+1), y(.tau.+2), . . . y(N).
Then, a correlation coefficient (CC) may be expressed as a function of
.tau. by:
##EQU1##
where x.sub.m, y.sub.m are the mean value of x(n) and y(n) and
##EQU2##
For negative values of .tau., a similar formula may be used. The lag of
maximal correlation coefficient is taken as the lag between the two
parameters.
By one or more of the above techniques and/or other techniques, values of
the delay between the PPG signals are calculated for a range of applied
cuff pressures from above the expected systolic blood pressure to below
the expected diastolic blood pressure values. Optionally, one or more
additional stage of smoothing may be applied, firstly to the cuff pressure
as a function of time so as to remove peaks due to the pulse, and secondly
to the time delay itself as a function of pressure. The time delay values
are also normalized by subtracting a baseline delay value corresponding to
the delay measured when substantially no pressure is applied to the cuff.
The techniques described above have been found experimentally to give very
similar results. By way of example, FIGS. 13 and 14 show the variation of
.DELTA.t with applied cuff pressure as determined by the techniques of
FIGS. 10 and 11, respectively.
Once this information is determined, the diastolic blood pressure is
identified according to the predefined non-zero value of delay measured
relative to said baseline value. Through extensive research with many
subjects, it has been found that the first diastolic blood pressure,
equivalent to that identified by the conventional auscultatory method,
consistently corresponds to a delay of between about 30 and about 40 ms,
and most preferably about 35 ms, above the baseline delay. Similarly, the
second diastolic blood pressure, equivalent to that identified by the
conventional auscultatory method, consistently corresponds to a delay of
between about 15 and about 25 ms, and most preferably about 20 ms, above
the baseline delay.
While the aforementioned values of the time delay are believed to be
optimal for derivation of the conventionally used diastolic blood pressure
values, it should be noted that other values may alternatively be used to
provide other meaningful indicators of diastolic blood pressure.
Additionally, measurement at a higher value of the delay may be used
together with a predetermined subtraction or addition of a pressure value
to obtain estimated blood pressure values corresponding to the
conventionally used values. In this context, it is noted that measurements
taken in the range between about 40 and about 60 ms, and in particular at
a value chosen at or near 50 ms, are advantageous due to the smoother form
of the results generally obtained in those ranges.
Turning now to FIG. 15, this illustrates the principle of a variant method
according to the present invention. In this case, instead of determining a
value of the time delay between the two pulses, a correlation function
applied directly to the two signals is used as an indication of the degree
of lag between them. The function may be as defined above with a constant
.tau.=0.
More specifically, as before, the method includes generating first and
second signals indicative, respectively, of cardiac induced pulsatile
variations in tissue blood volume in a first region and a second region of
the subject's body according to any of the devices described above. The
first and second signals are then processed to derive values of a measure
of correlation between pulses in the first signal and corresponding pulses
in second signal. These values are adjusted by addition or subtraction of
the difference between 1 and a baseline value corresponding to the
correlation coefficient when no pressure is applied to the cuff such that
the correlation coefficient takes a maximum value of 1. An example of the
resulting values as a function of time is shown in FIG. 15.
The diastolic blood pressure is then identified as the value of the
variable pressure corresponding to a value of the measure of correlation
substantially equal to a predefined proportion of the baseline value. This
predefined proportion is at least about 90%, i.e., 0.9 on the adjusted
plot. The actual values of the correlation coefficient determined
experimentally which are believed to provide accurate values corresponding
to conventional diastolic blood pressure values are about 0.915 for the
first diastolic pressure and about 0.955 for the second.
In an alternative implementation, the measured values may be scaled by the
base-line value rather than adjusted by addition or subtraction.
Turning now to FIG. 16, it will be noted that preferred implementations of
the device and method of the present invention also provide indications of
the systolic blood pressure. In principle, this can be achieved trivially
by sensing the cessation and recommencing of PPG pulses as the applied
cuff pressure goes above and returns below the systolic pressure. In
practice, however, the pulses emerge gradually from the background signal
noise in a manner which renders it difficult to identify a clear cut-off
for the systolic pressure.
To address this problem, the present invention preferably provides a method
to quantify the reappearance of the PPG signal. To this end, the PPG
pulses occurring before the cuff pressure increase are used as a
reference. A number of pulses, typically between about 4 and about 10
pulses, are recorded and their mean amplitude (or area) is stored. The
systolic blood pressure is then taken as the cuff pressure for which the
amplitude of PPG signals (or their area) is a predefined proportion of the
stored mean amplitude (or area). A proportion of about 8-10% has been
found sufficient to distinguish the signal from background noise while
providing sufficiently accurate results.
FIGS. 17A-17C illustrate the comparison between the present method (by use
of 0.1 of the amplitude) and sphygmomanometry which is the "gold
standard". The average and standard deviation of the difference between
the two methods are as follows:
______________________________________
mean difference
standard deviation
(mmHg) (mmHg)
______________________________________
Systolic -1.9 3.4
Diastolic -1.5 4.9
______________________________________
These compare favorably with the standards set by AAMI (US) and BSH (UK).
By way of example, the standards required by AAMI are a maximum mean
difference of .+-.5 mmHg and a maximum standard deviation of 8 mmHg. A
good accuracy has been achieved in reproducing the systolic and diastolic
arterial blood pressure values using the device and methods of the present
invention in a reasonable range of pressures.
While in the above embodiments the pressure application means has been
shown and described to be attached to a subject's arm, it should be
understood that it can just as well be attached to a subject's hand, leg,
ankle or foot, arms and legs being referred to collectively as limbs.
Similarly, it should be understood that the measurement probes of the
present invention may be attached to a subject's legs, feet or toes.
Fingers and toes are referred to collectively as digits.
It will be evident to those skilled in the art that the invention is not
limited to the details of the foregoing illustrated embodiments and that
the present invention may be embodied in other specific forms without
departing from the spirit or essential attributes thereof. The present
embodiments are therefore to be considered in all respects as illustrative
and not restrictive, the scope of the invention being indicated by the
appended claims rather than by the foregoing description, and all changes
which come within the meaning and range of equivalency of the claims are
therefore intended to be embraced therein.
Finally, turning to FIGS. 18 and 19, it should be noted that the devices of
the present invention facilitate implementation as a compact,
non-intrusive device which can be worn by a patient for continuous or
intermittent monitoring of blood pressure for extended periods without
interfering with other activities. To this end, the implementations of
FIGS. 18 and 19 preferably employ PPG sensors formed as rings which can be
worn non-intrusively on a finger of each hand. The main electronic control
and processing elements are preferably implemented as a separate box which
can conveniently be worn on a belt or the like.
In the implementation of FIG. 18, connections between the PPG sensors, the
cuff and the control box are made via wires passing up the arms. These may
be taped or otherwise attached to the skin beneath clothing in a manner
generally non-disruptive to a wide range of activities.
FIG. 19 shows a further option in which some, or preferably all, of the
components are associated via wireless connections, typically at radio
frequencies. The miniature transmitter and receiver components required
for such implementations are widely available, as will be clear to one
ordinarily skilled in the art.
It will be appreciated that the above descriptions are intended only to
serve as examples, and that other embodiments are possible within the
spirit and the scope of the present invention.
Top